1
Adjusting to multipolarity
in the World Bank: ducking and diving, wriggling and squirming Jakob Vestergaard and Robert Wade DIIS Working Paper 2011:24
W O R KI N G P A PE R
2
JAkoB VestergAArd
roBert H. WAde
DIIS Working Papers make available DIIS researchers’
and DIIS project partners’ work in progress towards proper publishing. They may include important documentation which is not necessarily published elsewhere. DIIS Working Papers are published under the responsibility of the author alone. DIIS Working Papers should not be quoted without the express permission of the author.
DIIS WORKING PAPER 2011:24
© The author and DIIS, Copenhagen 2011 Danish Institute for International Studies, DIIS Strandgade 56, DK-1401 Copenhagen, Denmark Ph: +45 32 69 87 87
Fax: +45 32 69 87 00 E-mail: [email protected] Web: www.diis.dk
Cover Design: Carsten Schiøler Layout: Ellen-Marie Bentsen
Printed in Denmark by Vesterkopi AS ISBN: 978-87-7605-470-0
Price: DKK 25.00 (VAT included) DIIS publications can be downloaded free of charge from www.diis.dk
CONtENtS
Introduction 5
1. The governance of the World Bank 7
Shareholding and voting power 7
The Bank’s governing bodies 8
2. Evolution of the voice reform agenda 10
The 2003 Background Paper 10
The 2007 Options Paper 12
3. First phase of voice reform 13
Increasing basic votes 13
Realignment of quota votes 14
Increasing the voice of African countries on the Board 14
4. The second phase of voice reform 14
The key components of the shareholding realignment 16
Main results of the second phase of voice reform 18
5. The voting power realignment in perspective 19
Modest changes 19
The country reclassification game 20
Making small changes appear generous 21
Voting power imbalances 22
6. Problems and solutions 23
Voting power 24
The composition of the Board under pressure 26
7. Conclusion 27
References 29
4
INtRODuCtION 1
The world is experiencing the biggest shift in the location of economic activity in centuries;
roughly speaking, from West to East. The shift is creating deep insecurity in the long domi- nant states, prompting them to cling to the power positions they attained as a result of their economic dominance in the decades after the Second World War. It is creating ambiguity in the rising economic powers about a new role in inter-state organizations, wanting a larger voice but also wary of new responsibilities. The lat- est battle ground is the International Monetary Fund when Dominique Strauss-Kahn stepped down as director. European states lobbied hard to retain Europe’s (France and Germany’s) long monopoly on the position, and succeeded with the appointment of former French finance minister Christine Lagarde.
This essay examines the efforts of the World Bank to translate the rising economic weight of some developing countries into a larger voice in the governance of the organization. In a speech in April 2010, World Bank president Robert Zoellick argued that the advent of “a new, fast-evolving multipolar world economy”
required fundamental reforms of the World Bank itself, including in the balance of power between developed countries and emerging countries Shortly after the speech the World Bank presented a set of allegedly far-reaching proposals on what it called “voice reform”, to be endorsed by its Board of Governors, the culmination of a process of negotiation begun years before. Voice reform had several compo- nents, of which the central and most conten- tious one was voting reform to give developing
For short essays on the main outcomes of the voice re- form process see Horton (200) and Lombardi (200).
2 Robert Zoellick (200).
and transition countries (DTCs) more voting power in the Bank’s governance.
The Governors approved the proposals at the 2010 Spring Meetings of the World Bank and International Monetary Fund (IMF).
The headlines went around the world that the World Bank had approved a major increase in the governance role of its borrowing member states, partly by redistributing voting power in their favour. The Bank had successfully trans- lated changes in economic weight in the world economy into changes in political weight with- in the organization.3
This sets our question: to what extent does the new distribution of votes bring the organi- zation more closely into line with the distribu- tion of economic weight in the world econo- my? Our answers are based on more than forty interviews with Bank staff and analysis of voice reform documents from the first “scoping”
paper in 2003 onwards.
We find, first, that the voice reform process accomplished a total shift of voting power of 4.59 percentage points from developed coun- tries to developing and transition countries (DTCs), increasing the share of DTCs from 42.60 % to 47.19 % and reducing the share of developed countries from 57.40 % to 52.81
%. But the shift was in fact more modest than these figures suggests because the DTC cat- egory was massaged to include several high- income countries which had no business to be there. With only low-income and middle- income countries included (using the Bank’s own categorization), the shift from high to low and middle-income countries was only 3.71 percentage points, taking the share of the latter from 34.67 % to 38.38 % while the high-in- come countries retained more than 60 %.
Much the same shift was agreed at the IMF a little later.
Second, relative to the alleged objective of rea- ligning voting power with the realities of the fast evolving “multipolar” world economy, the realignment was quite inadequate. So small were the shifts in voting power for the vast majority of countries that one exasperated observer described the process as a search for
“compromises at the third decimal point”. The upshot is that “voting power to GDP” ratios in the World Bank continue to vary widely from country to country, from 0.5 to 4, de- spite the oft-cited principle that voting power should “largely reflect economic weight” (so that each country’s ratio should be fairly close to 1). A number of small European countries and a few large DTCs have disproportionately large amounts of voting power, while several dynamic emerging market economies, includ- ing China, continue to be significantly under- represented. The eight-fold difference in the extent to which GDP translates into voting power undermines the normative legitimacy of the World Bank.
Third, despite repeated assurances to the contrary, low-income countries as a group (as distinct from middle-income countries) gained hardly any voting power. This reflects a pattern in which the interests of the low-income coun- tries were marginalized in the voice reform.
The culmination of this trend was the decision to make only a very small increase of “basic votes” (votes allocated equally to all countries), leaving the share of basic votes in total votes at only about half of what it was when the World Bank was established in 1944.
Fourth, the voice reform made no headway in reaching agreement on criteria for reallocat- ing votes in future (except for the agreement that shareholding reviews should be conduct- ed every five years). For example, it is unclear whether the next shareholding review in 2015 will take “voting power parity” between devel- oped countries as a group and DTCs as a group as the central objective, and whether and how
a country’s financial contributions to IDA (the soft-loan arm of the World Bank) should be recognized in its share of IBRD votes (IBRD being the main lending arm).
Fifth, the fact that all member countries have a veto over any decrease in their share of World Bank (IBRD) votes was and will be det- rimental to any process of adjustment of World Bank governance. The Articles of Agreement must be changed to remove this right of veto.
The voice reforms at the World Bank can be seen as its attempt to reconcile two differ- ent requirements for authority, requirements which all inter-state organizations have to balance: normative legitimacy, and interest- based demands of member states.4 Normative legitimacy depends on the extent to which the organization’s procedures are in line with normative principles, including agreed prin- ciples for the allocation of power within the organization. High normative legitimacy in- clines members to confer authority on the or- ganization – that is, to grant it the ability to promulgate rules that are implemented and followed. But members also confer authority on the organization to the extent that it serves their national interest. There is a tension: serv- ing the national interest of key member states may undercut normative legitimacy, by leading the organization to fudge its compliance with its principles (of representation, for example).
This essay shows how the World Bank has con- torted itself to satisfy three divergent sets of expectations about its allocation of governing power: (1) the normative principle that voting power (and “voice” more generally) should be proportional to economic weight in the world economy (plus some qualifiers to protect small and poor economies); (2) interest-based de- mands of the long dominant states, notably
4 See Koppell (2010, chapter 2); Wade (2002).
the United States and western Europe; and (3) interest-based demands of some of the newly emerging states. Of course, the art of states pressing interest-based demands is to cloak them in normative justifications.
Section 1 gives a brief overview of the Bank’s governance arrangements, and section 2 describes the evolution of the voice reform agenda in the Bank from 2003 to 2007. Sec- tion 3 examines the outcome of phase 1 of the voice reform, completed in 2008, and section 4 does the same for phase 2, completed in 2010. Section 5 makes an assessment of the voting power realignment in the light of the Bank’s avowed normative principles. Section 6 describes some of the problems left unre- solved for future shareholding reviews. Section 7 summarizes the main findings and recom- mendations.
1. thE GOvERNANCE Of thE WORlD BANK
The three main components of the World Bank Group (WBG) are the International Bank for Reconstruction and Development (IBRD), the International Development Association (IDA) and the International Finance Corpo- ration (IFC).6 While formally there are three separate Boards for IBRD, IDA and IFC, the same people are on the Boards of each (with different voting power depending on whether
Key references in the scholarly debate on governance reforms of the World Bank include Birdsall (2006), Buira (200), Kapur (2002), Phillips (2009), Weaver and Leiteritz (200), and Woods (2008a, 2008b).
The World Bank Group is completed by three additional affiliate organizations:, the Multilateral Investment Guaran- tee Agency (MIGA) and the International Centre for Settle- ment of Investment Disputes (ICSID).
the subject matter at hand concerns the IBRD, IDA or IFC).
The voice reform focused on the main lend- ing arm, the IBRD, because while sharehold- ing differs for IBRD, IDA and IFC, it is IBRD shareholding that determines the structure of all three Boards (DC 2010a: 3). This stems from the fact that, legally speaking, country constituencies are formed and Executive Di- rectors of the Boards are elected on the basis of IBRD shareholding.
Shareholding and voting power
The IBRD was established in 1944 as the origi- nal organization of the World Bank Group.7 The shareholding of its 187 member countries is comprised of two elements: basic votes and quota votes. Basic votes are allocated to all mem- bers in the same amount, and quota votes are allocated in proportion to shares subscribed to.
This combined system of basic votes and quota votes was a compromise between two factions at the original Bretton Woods conference, “re- spectively preferring a one member–one vote system and voting based purely on the size of each country’s economy” (Woodward 2007:
1). Introduced at the founding of the IBRD, basic votes were to ensure voting power for the smaller and poorer member countries. The Ar- ticles of Agreement stipulate that the number of basic votes shall be 250 per member, and this number has been held constant ever since
The IBRD aims to “reduce poverty in middle-income and creditworthy poorer countries by promoting sustain- able development through loans, guarantees, risk manage- ment products, and analytical and advisory services” (DC 200a).
1944. 8 Over the years the share of basic votes in total votes has eroded to just 2.8% from the initial level of more than 10%, as the alloca- tion of quota votes increased.9
On top of the 250 basic votes, each mem- ber country has one additional vote for each share of stock held (IBRD Article V, section 3a). One share gives one quota vote. For each share of stock, a certain proportion (currently about 6%) must take the form of paid-in capi- tal, while the remainder is “callable”. Although the notion of shareholding might imply so, there is no market for IBRD shares.10 Instead, IBRD shares are allotted to member countries in proportion to their “relative position” in the world economy (at least in principle). In the words of the World Bank:
The fundamental principle underlying the allocation of shares of the IBRD’s capital stock to its members is that mem- bers’ subscriptions should reflect their relative position in the world economy, subject to the right of each member to maintain its existing pro rata share in the capital on the occasion of any increase in the authorized capital (pre-emptive right). (DC 2003a: 11-12).
8 In 979 all member countries were invited to subscribe to an additional 20 “membership shares”, which would have amounted to a doubling of basic votes (DC 2003a: 8;
DC 2008a). Not all subscribed to these membership shares, and therefore the sum of basic votes continue to be stipu- lated at 20.
New quota votes were allocated in the context of a number of selective capital increases by which some coun- tries paid in more capital and were allotted additional quota shares, with no accompanying adjustment of basic votes.
0 Countries may choose to “subscribe” to less than their allotted shares, in which case the unallotted shares go into a pool. The country then makes lower capital contributions and receives a lower share of total votes.
Historically, the World Bank has operational- ised the criterion of proportionality between shares and weight in the global economy by establishing a close link between IBRD share- holding and IMF quotas. This is odd, because the formula for IMF quotas gives only 50%
weight to GDP, the other components being openness (30%), economic variability (15%) and international reserves (5%) – components which are less relevant to the World Bank’s mandate than to the IMF’s. The voting power reform which was agreed in 2010 abandoned the close link to the IMF quota formula. It was based on a quota “framework” developed exclusively for World Bank (IBRD) share- holding, with only indirect reference to IMF quota.11 Taken at face value, the 2010 World Bank voting power framework gave stronger weight to GDP (75%) than in the IMF formu- la (50%), suggesting a close and strengthened link between GDP and voting power. In fact, however, voting power to GDP ratios in the Bank vary almost inexplicably, as we shall see.
the Bank’s governing bodies
All member countries have direct representa- tion as members of the Board of Governors, at the level of ministers. It convenes twice a year, once at the Spring Meetings of the World Bank and the IMF, and once during the An- nual Meetings in the autumn. The role of the Board of Governors is limited, however. It del- egates its authority to a subset of its members, which constitutes the Development Commit- tee. But deliberation and negotiation amongst the member countries mainly take place in and through the Executive Board of Directors
11 Confusingly, however, the World Bank continues to sug- gest otherwise at its website, explaining that ‘the quota as- signed by the Fund is used to determine the number of shares allotted to each new member country of the Bank’
(WB 20).
(EBD), a resident body comprised of civil serv- ants, based in Washington. The EBD has over- all responsibility for the general operations of the Bank and exercises all the powers delegated to it by the Board of Governors, which includes both executive and oversight functions.12
At first the EBD consisted of 12 Executive Directors, as prescribed in the IBRD Articles of Agreement (Article V, Section 4b). The five largest shareholders in the Bank were granted the right to appoint their own Executive Di- rector, while the other seven were elected Ex- ecutive Directors, based on country constitu- encies. Over the years, the total number of Executive Directors has increased to 25. Most of the increase occurred before the 1990s. Since then, three new seats has been created: Russia got a seat of its own in 1992, a new constitu- ency was formed around Switzerland also in 1992, and a third seat was allocated to Africa in November 2010 as part of the voice reform.
There are now eight single-country seats (US, UK, France, Germany, Japan, China, Russia, Saudi Arabia), and 17 multiple country con- stituency seats (see the Appendix for overview of all constituencies). Of the multiple coun-
12 The day-to-day operations of the Board include delib- erating on proposals made by Bank management on IBRD loans and guarantees, IDA credits and grants, IFC invest- ments and policies that “impact on the World Bank’s gen- eral operations” (WB 20). In addition to its executive functions, the Board has oversight functions, and two World Bank bodies report directly to the Board to help it perform this role: the Independent Evaluation Group (IEG) and the Inspection Panel. Although it is in principle within the pow- ers of the Board to hire and fire the President, in fact he is appointed by the US and he is primarily accountable to the President of the United States and the US Congress. This, of course, has been a subject of contestation, not least in the context of voice reform deliberations, and numerous are declarations that commit the Executive Board of Directors, the Board of Governors and the Development Committee to select future Presidents of the World Bank on the basis of an “open, transparent and merit-based process”. So far the declarations have had as much effect as pushing on a piece of string.
try constituencies many are so-called “mixed constituencies”, where developed countries and DTCs share a seat. Spain, for instance, currently holds the Executive Directorship of a country constituency that includes Mexico, Costa Rica and Venezuela, among others. The Development Committee mirrors the compo- sition of the EBD but at the political rather than civil service level.
The voting system is based on the share- holding of the member countries that have ap- pointed or elected a given Executive Director.
While the same persons are on the Board of Executive Directors of the IBRD, IDA and the IFC, their respective voting power depends on which of these three WBG bodies a given vote is cast for, since countries’ relative sharehold- ing is not the same for each of the three bodies.
Most decisions require a simple majority, al- though there are important exceptions to this rule. Special majorities are required for issues such as capital increases and amendment of the Articles of Agreement. Amendment of Articles requires approval by the Board of Governors, with support from at least 60% of member countries and at least 85% of total voting power (DC 2007b: Annex II).
The latter criterion is what gives the US a veto on constitutional changes. Given that the US has just over 15% of total voting power, no amendment of the Articles can be decid- ed without US agreement. 13 Increases in the Bank’s capital also require a special majority, although here only a 75% majority of voting power applies (DC 2003b: 5). It is important
Special majorities in both the World Bank and the IMF have changed over time to ensure that the US kept its veto power even as its share of voting power declined. In the case of the IMF, “a special majority of 7% of votes was required when US voting power was just over 2%. That special majority requirement is now 8%, retaining a US veto power even though US voting power has slipped to 7%” (Woods, 2008b). See also Woods (2008a).
10
to note that in the context of an increase of the Bank’s capital, each and every member coun- try has a right to “subscribe to a proportion- ate share of the increase” (DC 2007b: 5). This in effect means that no member country can have its share of total shares reduced without its concurrence, a principle known as “pre- emptive rights” (IBRD Article II, sections 2b, 3b and 3c). The implication is that, since any realignment of voting power requires a selective capital increase (with those who gain voting share contributing more capital than others who loose), voting power reform can only be undertaken if all 187 member countries agree unanimously. This is not a misprint.
2. EvOlutION Of thE vOICE REfORm AGENDA
The voice reform process originated in the Monterrey Consensus, articulated at the United Nations International Conference on Financing for Development in Monterrey on 22 March 2002. While the main elements of the Monterrey Consensus were agreements on such issues as debt relief, development aid and fighting corruption, the communiqué includ- ed an important commitment to enhance the voice and participation of developing countries in multilateral organizations:
We stress the need to broaden and strengthen the participation of devel- oping countries and countries with economies in transition in international economic decision-making and norm- setting … A first priority is to find prag- matic and innovative ways to further en- hance … effective participation … and thereby to strengthen the international dialogue and the work of [multilateral organizations] as they address the de- velopment needs and concerns of these countries (UN 2003: 20)
For several years after the Monterrey Consen- sus deliberations on voice reform in the World Bank moved at a snail’s pace. No one was keen to lead, especially not when the impetus came from the United Nations, a body not held in high esteem within the World Bank.
The first background report was prepared for the 2003 Spring Meetings, and the com- ing years saw a number of progress reports and further background reports prepared for Spring and Annual Meetings, culminating in an Options Paper for the 2007 Spring Meetings (DC 2007b).14
the 2003 Background Paper
In response to the Monterrey Consensus, the Development Committee requested the World Bank and the IMF to prepare a joint back- ground paper to “facilitate consideration, at its Spring 2003 meeting, of ways of broadening and strengthening the voice and participation of developing countries and countries with economies in transition” in the two organiza- tions (DC 2003a: 1). The Background Paper noted that a ‘broad degree of consensus’ would be required for voice reform to succeed, and then proceeded to outline the key issues and the possible avenues to pursue. It identified three main issues.
First, the relative voting power of member countries, and particularly the question of the extent to which some countries might be said to be “over-represented” and others
“under-represented”.
Second, the challenge of ensuring regional balance: the paper noted that “significant
14 In the interim period, voice reform was on the agenda of the Development Committee three times – in Fall 200, Fall 2004 and Spring 200.
•
•
11
changes in the regional composition of the Boards to strengthen developing country participation would require understand- ings among the membership on what regions are ‘under’- or ‘over-represented’”
(DC 2003a: 2, emphasis added).
Third, the problems of ensuring voice and participation for countries that are mem- bers of very large country constituencies, given the complexity of coordination in these constituencies.15 The two largest constituencies have 18 and 23 member countries, respectively.)
Before discussing possible options for address- ing these issues, the Background Paper identi- fied two key issues on which there was such broad agreement as to warrant no more con- sideration:
the constituency-based system of represen- tation, and
the principle that voting power should “in large measure reflect the relative import- ance of member countries in the global economy” (DC 2003a: 3).
It is important to highlight these two alleged areas of broad consensus, since subsequent developments show much dissensus around them.
The Background Paper divided its consid- eration of options into two main categories:
proposals to enhance voice and proposals to enhance voting power:
1 This problem is further aggravated by the severe imbal- ances in the resources made available for different country constituencies by the governments of their member coun- tries, notably the modest resources available for most Ex- ecutive Directors representing developing countries.
•
•
•
Proposals to enhance voice
First, the paper discusses several administra- tive fixes to the problems of large multi-coun- try constituencies. It states that support for these constituencies might take many different forms, ranging from the provision of techno- logical assistance to facilitate communication with capitals (video conferencing, etc.) and es- tablishment of a trust fund to support research and analysis for some multi-country constitu- encies, to supporting the employment of ad- ditional assistants and the addition of a second Alternate Executive Director.
Second, it mentions more politically or le- gally demanding measures to enhance the voice of developing countries, not least the possibil- ity of increasing the number of Board seats so as to reduce the number of member countries in the largest constituencies, and of reviewing the regional composition of the Boards. “A re- duction in the number of Executive Directors appointed or elected by industrial countries, combined with a rearrangement to reduce the number of countries in the largest con- stituencies, could be seen as proportionally strengthening the voice of developing country Directors in the Boards”, the paper notes (DC 2003a: 6).
With regard to this latter option the Back- ground Paper also notes that such “significant changes” would “raise a set of complex issues”
and would require “broad-based political con- sensus among the membership” (ibid.). More specifically, an amendment of the Articles of Agreement would be required in order to ad- just the rights of member countries standing to lose their entitlement to appoint their own Executive Director.
Proposals to enhance voting power
The Background Paper acknowledged that the
“most straightforward dimension” of voice and participation is voting power on the Boards of
12
the Bank and the Fund (DC 2003a: 1). Nev- ertheless, it gives considerably less attention to the options in this area than on broader aspects of voice and participation: of the paper’s nine pages only one is devoted to “possible avenues for enhancing voting strength” (DC 2003: 8).
Further, although increasing developing coun- tries’ shareholding is recognized to be “the most direct way” of enhancing their voting power, this option is mentioned only to reject it (DC 2003a: 8). “There is not at present sufficient support”, the Paper declares, for initiatives
“that might lead to an increase in the overall voting share of developing countries” (ibid.).
The paper instead directs attention to the other main mechanism for enhancing the vot- ing power of developing countries, namely a uniform increase in member countries’ basic votes (an equal number to each country would raise the share for developing countries propor- tionately more). But the brief discussion of this option also ends on a pessimistic note, with the observation that this proposal had “been made from time to time, but lacked wide support”
and that an increase in “basic votes requires an amendment of the Articles of Agreement”
(ibid.). It also discusses a third option, namely to increase the use of special majorities for spe- cific types of decisions. “It has been suggested that requiring a special majority of 70-85%
of votes on critical decisions could give addi- tional assurances that the voice of developing countries will be heard and considered” (DC 2003a: 9). However, such an increased use of special majorities “would be likely to favor the status quo”, it says, “and it is not clear that it would, in practice, have the effect of increasing developing country voice” (ibid.).
the 2007 Options Paper
Those were the lines of thinking in the ini- tiating paper of 2003. Now fast forward to 2007 when the Bank produced the “Options
Paper”.16 The paper notes that between 2003 and 2007 little progress had been made. Why?
“Limited debate on Voice and the overall lack of progress … are due to the lack of political consensus” on issues such as IBRD’s voting structure; potential changes in IBRD’s capital stock; and the composition of the Board of Executive Directors (ibid.).17
To push things forward, the Options Paper proposed a two phase program for voice re- form:
The first phase should move rapidly ahead with “an initial package of options which holds the promise to generate consensus and help build momentum” in areas such as appointment of more DTC nationals in senior management positions, procedures for selection of the Bank’s President and for Board effectiveness (DC 2007b: 17).
The second phase would then “address the more challenging structural options for which a political consensus can be achieved as early as possible”, such as a possible increase in basic votes and a selec- tive capital increase on the part of those countries whose voting share was to be increased (DC 2007b: 17-18).
Although the (politically-constituted) Devel- opment Committee had not discussed voice reform much since the 2003 Annual Meet- ings, extensive deliberations had been going on among the (civil service) Executive Direc-
1 The process was by then separate for the Bank and the Fund, and had been for a couple of years.
1 Progress is noted in one area, namely capacity building.
The paper mentions two examples of voice enhancing ca- pacity building: the establishment of an analytical trust fund
“to provide sub-Saharan EDs [executive directors] with independent technical research support” and a multi-year secondment program for DTC officials in the Bank (DC 2007b: 3)
•
•
1
tors of the Board in the interim. Their debates inform the inventory presented in the 2007 Options Paper. It is noteworthy that out of the ten main options considered in the paper, nine relate directly to IBRD voting structure, IBRD capital stock or composition of the Board – i.e.
those areas of voice reform that were treated only cursorily in the 2003 paper. That these areas of reform had now moved centre stage indicates progress of sorts.
However, most of the options required consensus or a large special majority, and the Options Paper reported that the voice reform agenda was “an issue on which agreement on a way forward has been elusive with no signifi- cant progress made” (DC 2007b: 16, emphasis added). The Paper then identified a set of ‘con- cerns’ with respect to which agreement was necessary:
The need to realign the shareholdings and voting rights of member countries with their changed position in the global economy.
The need to take into account donors’
contributions to IDA and to overall ODA (official development assistance), including the funding of World Bank trust funds.
The need to prevent or at least contain the erosion of the position of smaller countries which, although they have a small share of the global economy, represent a significant focus of the Bank’s work.
The fact that four years of deliberations of member countries, mainly at the level of Exec- utive Directors, had led to little agreement on any of them underlines the inherent difficulties of reforming the governance arrangements of the World Bank.
•
•
•
3. fIRSt PhASE Of vOICE REfORm At the Spring Meetings of 2008 the Develop- ment Committee encouraged “the Bank to advance work on all aspects of voice and par- ticipation, keeping in mind the distinct nature of the Bank’s development mandate, and the importance of enhancing voice and participa- tion for all developing and transition countries in the World Bank Group” (DC 2008a). Later that year, at the Bank’s Annual Meeting, the De- velopment Committee endorsed the first pack- age of voice reforms (DC 2008b), as outlined in what was called the “Background Paper by Board and Management” (DC 2008c).18
The Background Paper of 2008 summarized for each reform area a wide range of options, and concluded by identifying and recommend- ing a much narrower set of concrete decisions (but without giving the grounds on which one option was chosen over others). It identified three main areas of voice reform:
An increase in basic votes
A realignment of quota votes to better reflect countries’ relative weight in global economy
The addition of a third Executive Director for the African countries 19
Increasing basic votes
The main element of the first phase of the voice reform agreed at the 2008 Annual Meetings was an increase in basic votes, for the explicit pur- pose of benefiting the poorest developing coun- tries. The Background Paper of 2008 discussed three options:
8 The list of concrete decisions recommended can be found in DC (2008c, section 66a, p. 20-2).
9 Other areas of reform, beyond the scope of this paper, were voting power for DTCs (Part 2 countries) in IDA and voice reform for the IFC.
••
•
14
First, the Doubling Option, by which basic votes would be doubled to reach 500 per member, or 5.55% of total votes.
Second, the Tripling Option, by which basic votes would increase to 750 per member, taking basic votes to 8.1% of total votes.
Third, the ‘Original level’ Option, by which basic votes would be reset at the level originally agreed in 1944, i.e. at 10.78% of total votes.
Eventually, the Development Committee reached agreement on the Doubling Option, the most conservative. The decision to increase basic votes by 250 per member was accom- panied by a decision to maintain the share of basic votes in total votes at 5.55%, at least, in the future. In terms of the relative voting power of developed countries and DTCs, the doubling of basic votes was predicted to result in a shift of 1.2 percentage points from the former to the latter.
Realignment of quota votes
The Background Paper discussed three main options by which a realignment of quota votes might be achieved: a selective capital increase, an allocation of unallocated shares, and a share exchange (DC 2008c: 8). But it ended up by proposing further deliberation, in the form of a review that “would lead to a subsequent significant realignment” of quota votes for all member countries so as to “further enhance the Voice of DTC members” and “address the concept, advocated by some members, of mov- ing over time towards equitable voting power between developed and developing members”
(DC 2008c: 10). Note the odd language. The proposal is not for the review to ‘move towards equity’ but to ‘address the concept of moving to- wards equity’. This phrasing reflects deep-seated disagreement between developed and develop- ing countries about the overall target of the
•
•
•
voice reform process. The Board of Governors accepted the proposal.
Increasing the voice of African countries on the Board
The most progressive result of the first phase of the voice reform process was the decision to expand the Executive Board of Directors so as to allow a third African seat. Given both exter- nal and internal pressure to reduce rather than increase the number of Executive Directors on the Board, the successful negotiation of a third African seat by the Executive Directors was all the more significant. 20
4. thE SECOND PhASE Of vOICE REfORm
The second phase of the voice reform was originally planned for the 2011 Spring Meet- ings. However, the first phase ended at the An- nual Meetings in September 2008, just as the world economy went into meltdown. Soon afterwards President George Bush convened the G20 Leaders summit for the first time, which took upon itself to issue instructions to the Bretton Woods organizations about their reforms. Voice reform entered the zeitgeist as a top priority, not something that those who stood to loose could keep dragging their feet on. So the 2009 Spring Meetings in Washing- ton agreed to accelerate the process. “The global economy has deteriorated dramatically since
20 The decision to grant a third seat to the African coun- tries – championed by the Nordics – was initially not widely supported on the Board. In fact, at first, few Executive Di- rectors took the idea seriously. In 2009 the Zedillo report – commissioned by President Zoellick – urged a reduction of seats from 24 to 20 (Zedillo, 2009). For a brief commen- tary on the main findings of the Zedillo Commission see Martinez-Diaz (2009).
1
our last meeting”, the communiqué noted, with “especially serious consequences” for de- veloping countries, since the “[h]ard-earned progress towards the Millennium Develop- ment Goals (MDGs)” was now considered to be “in jeopardy” (DC 2007a). The immediate impact of the crisis and the need to launch a
“strong multilateral response” to it motivated member countries to agree on significantly ac- celerating the voice reform process. “We agree to accelerate our work on the second phase of the reform”, the communiqué said, “with a view to reaching agreement by the 2010 Spring Meetings” (ibid.).
Now the big obstacle could no longer be avoided: the second voice reform was mainly about transferring voting power from devel- oped countries to developing countries in gen- eral and dynamic emerging market economies in particular. The target of the IBRD voting power realignment came from the G20 lead- ers at their summit in Pittsburgh in September 2009. The communiqué said:
We stressed the importance of adopting a dynamic formula at the World Bank which primarily reflects countries’ evolv- ing economic weight and the World Bank’s development mission, and that generates an increase of at least 3% of voting power for developing and transi- tion countries, to the benefit of under- represented countries. While recogniz- ing that over-represented countries will make a contribution, it will be impor- tant to protect the voting power of the smallest poor countries (G20 2009, em- phasis added).21
21 Note that the G20’s instruction for a shift of at least percentage points was to be on top of the already achieved shift in the first phase of the voice reform.
During the Fall of 2009 a process of negotia- tion began, based on a range of options and scenarios prepared by Bank Management (DC 2009). The key issues were: which indicator to use for economic weight and which criteria other than economic weight to include.
The negotiations considered using either the IMF quota formula to measure economic weight, or one of several “GDP blends” (60/40, 50/50, 40/60 and 30/70, where the figures refer to the percentage weight of GDP at mar- ket exchange rates and at purchasing power parity [PPP]). Eventually, the Board decided to use GDP as the benchmark for economic weight, not the IMF quota (thus ending the Bank’s long use of the IMF quota formula to make the first-cut allocation of its own voting shares). But deciding which GDP blend to use was not simple. Generally, developed countries prefer economic weight to be based on GDP at market exchange rates, whereas DTCs want it to be based on GDP at PPP – for the obvi- ous reason that the GDP of DTCs tends to be significantly higher at purchasing power parity than at market exchange rates, and vice versa for developed countries. In the end the Board decided that the IBRD shareholding realign- ment should follow the criterion adopted in the 2008 IMF Quota and Voice Reform, a weighted average of GDP at market values (60%) and GDP at purchasing power parity (40%). 22 This was the most conservative of the options.
The DTCs initially said that no developing country should lose voting power in this sec- ond phase of the voice reform and that addi- tional criteria beyond GDP had to be included to ensure that outcome. Developed countries, on the other hand, felt that several DTCs (such
22 This “GDP blend” is referred to in the remainder of this essay as “GDP (60/40)”.
1
as Saudi Arabia) were significantly over-repre- sented and should contribute to achieving the
“at least 3 percentage points” target by taking cuts in their voting power. Meanwhile, many developed countries emphasised that generous contributions to IDA should be recognized in IBRD shareholding, to the dissatisfaction of many DTCs; IDA and IBRD are different organizations, so why should contributions to one give shareholding in the other? Only on one thing did all parties seem to agree: the vot- ing power of the low-income countries should be maintained or expanded.
the key components of the shareholding realignment
Initially the aim was to arrive at a quota for- mula that would determine the shareholding of each member country. During the course of the negotiations it became clear that it would be necessary to settle for something less ambitious but more amenable to political compromise: a quota framework.
The quota framework set out overall princi- ples in a format that resembled a formula but wasn’t called one (especially because a “formula”
implied permanence and would be taken as the starting point for the 2015 negotiations).
The Executive Directors’ negotiations over the framework went on interminably, involving sev- eral iterations of ‘reverse engineering’ to get a quota framework that would yield a result po- litically acceptable to all parties.
The agreed quota framework set out three factors for determining quota votes: economic weight as measured by GDP (60/40), to receive a weighting of 75%; past and future contribu- tions to IDA (20%); and “contributions to de- velopment”, a measure of a country’s history of Bank borrowing (5%).
The GDP component
The paper that depicts the main elements of the voice reform claims that a realignment that simply brought ‘under-represented’ members’
shareholding up to their share of the global economy (based on GDP60/40) “would yield only a 1.3% increase in DTC voting power for Phase 2 reforms” (DC 2010a: 6). To achieve the “desired net increase of at least 3 per cent” a number of “adjustments” therefore would need to be adopted (ibid.):
First, only developed countries whose quota votes are below 90% of their calculated economic weights are eligible to take up ad- ditional shares to reach this 90% threshold.
Second, all DTCs whose quota votes are below their calculated economic weights are eligible to take up additional shares. 23 Moreover, the proposed voting power realign- ment was conditioned upon the voluntary forbearance of a number of under-represented countries including, most notably, China, Ger- many and the US.24
2 The final element of the GDP component of the IBRD quota framework was a so-called “PPP booster” which, modelled after a similar component in the 2008 IMF Quota and Voice Reform, gave countries whose “PPP-based weight in the world economy” was “0% or more above their IBRD shareholding a total increase in shareholding percentage of at least 10%” (DC 2010a: 7). Countries eligible for the PPP booster included Egypt, India, Indonesia and Uganda.
24 Quite a few European countries were identified as under-represented, including Germany, Greece, Italy, Ireland, Poland, Portugal, Spain and Turkey (DC 2010a: 24). Some of these countries chose to forego the increased shareholding they were entitled to (Germany, Greece, Portugal, Spain), while others did not (Italy, Ireland, Poland and Turkey). There were also DTCs in the list of countries identified as under- represented (Brazil, China, India, Indonesia, Korea, Mexico, Thailand, Vietnam etc), but here only China chose to forego its entitlement. The US was the sixth and final country that joined the small club of countries that chose to forego their entitlement to increased shareholding.
•
•
1
The IDA component
Historically, contributions to IDA had been recognised in terms of increased IBRD share- holding on an ad hoc basis for countries whose contributions were deemed particularly gener- ous. There had not, however, been established
“a rule or mechanism that takes regular IDA contributions of all IDA donors into account in IBRD shareholding” (DC 2009: 4-5). The main IDA donor countries were strongly in fa- vour of such a mechanism, whereas new IDA donors were more interested in a mechanism that would incentivize future IDA commit- ments. Eventually, the agreed IDA component recognized both past (actual) IDA contribu- tions and future (promised) IDA contribu- tions. 25
The framework for recognising IDA con- tributions ended up having two main dimen- sions, with two key components in each:
Recognition of past IDA contributions:
(1) Countries that meet criteria for recog- nition of contributions to the 13th through 15th replenishments of IDA are granted a 2% increase, while (2) countries that meet criteria for ‘historical’ IDA contribu- tion (pre-IDA 13) are granted a 1-1.5%
increase. Countries that meet both IDA13- 15 and historical IDA contribution criteria are granted a 3-3.5% increase.
Recognition of future IDA contributions:
(1) Current IDA donors are allocated shares in order to maintain their voting power if they increase their IDA16 contri- bution (agreed by end of 2010) by at least 50% over their IDA15 contribution, while (2) new IDA donors are allocated shares so
2 The 200 Options Paper speaks explicitly of the objec- tive of realigning IBRD shareholding so as to provide “in- centives for contributions to future IDA replenishments by current and new donors” (DC 200: , 2).
•
•
as to maintain their voting power if they contribute to IDA16 “at their notional IDA burden share” (DC 2010a: 5).
Not all were satisfied with this model for IDA recognition. The main IDA donors in recent decades – including Denmark, France, Ger- many, Norway, Sweden and the UK – com- plained that past IDA contributions carried far too little weight relative to pledges for future IDA contributions.26 Other critics complained that the future IDA contributions component was not a general mechanism to incentivize member countries to make IDA contributions but a mechanism to preserve the voting power of certain powerful DTCs including Russia, Saudi Arabia and Kuwait (because future IDA contributions were only recognised for those DTCs that would otherwise stand to lose vot- ing power). Without this provision these coun- tries might have been tempted to use their pre-emptive rights to block the voice reform, and their top-most leaders made it abundantly clear to Zoellick – as in haranguing phone calls -- that they would do so.
The third and final element of the quota framework, “development contributions”, was a mechanism to recognize “some of the many ways in which DTCs and their specific devel- opment experiences contribute to the World Bank Group” (DC 2010a: 9).27 In terms of real politik, this component was included to prevent increased voting power for dynamic emerging market economies from eroding the relative voting power of low-income countries.
2 Some observers lamented that “one dollar promised is given 400 times more weight than one dollar actually given”.
27 For further details see DC (2010a: 9-10).
1
main results of the second phase of voice reform
Overall, the main result of the voice reform process was a transfer of voting power from the US, Japan and a number of European coun- tries to dynamic emerging market economies (EMEs), especially China. Table 1 lists the 20 countries that gained most voting power and the 20 countries that yielded most voting power. It shows that the five countries that gave up most voting power were the World Bank’s five largest shareholders prior to the voice re- form, namely Japan, France, the UK, the US and Germany, while the main recipients were
China, South Korea, Turkey and Mexico. As a direct result of the voice reform China has now moved up to become the third largest share- holder after the US and Japan.28
2 With regard to the substantial increase of voting power for China, it should be stressed that China gave up as much as half of its entitlement (as measured by economic weight).
The Chinese said that they did so because otherwise other DTCs would have gained very little; while other observers suggested that Chinese forebearance reflected hesitation to move into high responsibility in the World Bank.
Table . Main results of voting power reform
Countries that increased their voting power
(percentage points) Countries that reduced their voting power (percentage points)
1 China (1.64) Japan (-1.01)
2 South Korea (0.58) France (-0.55)
3 Turkey (0.55) United Kingdom (-0.55)
4 Mexico (0.50) United States (-0.51)
5 Singapore (0.24) Germany (-0.48)
6 Greece (0.21) Canada (-0.35)
7 Brazil (0.17) Netherlands (-0.29)
8 India (0.13) Belgium (-0.23)
9 Vietnam (0.12) Switzerland (-0.20)
10 Spain (0.11) Australia (-0.19)
11 United Arab Emirates (0.09) Venezuela (-0.16)
12 Thailand (0.08) Italy (-0.14)
13 El Salvador (0.05) Nigeria (-0.10)
14 Costa Rica (0.05) Denmark (-0.09)
15 Romania (0.05) South Africa (-0.09)
16 Poland (0.04) Sweden (-0.09)
17 Indonesia (0.04) Ukraine (-0.09)
18 Tunisia (0.04) Algeria (-0.08)
19 Sudan (0.04) Austria (-0.07)
20 Panama (0.03) Pakistan (-0.07)
Source: DC 2010a.
Table 2. Regional profile of the voting power reallocation
Main receivers Main givers
Africa None None
Asia China, South Korea, Singapore, India,
Vietnam Japan
Europe Turkey, Greece, Spain France, UK, Germany, Netherlands, Belgium, Switzerland
Americas + Mexico, Brazil US, Canada, Australia
1
In terms of the distribution of increases and de- creases in voting power across the world’s four main regions, a majority of the major givers are from Europe (six of the top 10) while half of the receivers are from Asia (5 of the top 10).
See table 2, which lists the top 10 countries in these two categories by region.29
While the main focus of the voice reform was on IBRD shareholding, which determines the structure of the Board of Executive Direc- tors, the reform package also included vot- ing power realignments for shareholding in the IFC and IDA. The overall shifts of vot- ing power from developed countries to DTCs agreed upon in the course of the two phases of the voice reform process for each of these three Bank institutions, may be schematically represented as follows:
Table . DTC share of voting power in IBRD, IDA and IFC
Before Phase 1
(% increase) Phase 2
(% increase) Phase 1 + 2
(% increase) After
IBRD 42.60 1.46 3.13 4.59 (%)47.19
IDA 40.1 5.49 5.49 45.59 0
IFC 33.41 6.07 6.07 39.48
Source: DC 2010a: 14, 21 & 32.
Pre-Phase 1 Post-
Phase 1 Post-
Phase 2 Net change
(phase 1) Net change
(phase 2) Total net change (1+2)
LICs 3.45 3.94 3.84 0.49 -0.10 0.39
MICs 31.22 32.08 34.54 0.86 2.46 3.32
LICs + MICs 34.67 36.02 38.38 1.35 2.36 3.71
HICs 65.33 63.98 61.62 -1.35 -2.36 -3.71
2 This categorization of countries is based on UN sta- tistics, which divide the world in five regions: Africa, Asia, the Americas, Europe and Oceania. Oceania consists of Australia, New Zealand, Norfolk Island and three groups of island states: Melanesia, Polynesia and Micronesia. To get four regions we split Oceania and put Australasia (Australia, New Zeland, and Melanesia) with the Americas to make
“Americas+”; and put Polynesia and Micronesia in the Asia region
. thE vOtING POWER
REAlIGNmENt IN PERSPECtIvE modest changes
The voice reform was crafted so that it could be presented as taking the Bank half way towards the voting power parity objective wanted by developing countries: the three percentage point shift from developed countries to DTCs raised the DTC share of voting power from about 44% before the voice reform to about 47%. 3031
Here it is important to go behind the head- lines, and in particular, behind the category of
“developing and transition countries” (DTCs).
The bottom line is that by using the aggregate category of DTCs and by inserting in this cat-
0 DTCs share in IDA shareholding may potentially in- crease a further 2. percentage points. “If all available IDA subscriptions will be taken up, Part 2 shareholding in IDA could increase to 4. per cent” (DC 2010a: 14).
1 Net changes (last three columns in the table) are given in percentage points.
Table 4. The two phases of the IBRD voting power realignment (shareholding in %)
20
egory several high-income countries which are not borrowers from the Bank, the architects of the voice reform made it look as though a sig- nificantly bigger shift towards low- and middle- income countries had been effected than was actually the case. To see this, consider table 4, which summarizes the phases of the voting power realignment and its net effect on low and middle-income countries:
The high-income countries managed to limit their losses to the point where they remain above 60%: pre-reform they had 65.3%, now they have 61.6%. In other words, if parity is taken to refer to low and middle-income coun- tries on the one hand, and high-income coun- tries on the other hand – as would be natural in the case of the World Bank, which normally uses this three-fold income categorization – parity remains a long way off. The 3 percent- age point shift was only roughly one fifth of what would have been required to achieve par- ity in these terms (namely, a 15.33 percentage point shift, taking the aggregate share of LICs and MICs from 34.67 % to 50%).
Note also, with respect to the scale of the voting power reform for individual member countries, that only 22 of the 187 member countries were subject to a change of voting power of more than 0.1 percentage points, only 8 countries to a change of more than 0.5 percentage points, and only two coun- tries (China and Japan) to a change of more than one percentage point (see table 1). Voting power changes of this magnitude are a strik- ingly modest outcome of intense and costly deliberations in the Board over the course of almost a decade.
The country reclassification game Why did the Bank use the developed versus DTC categorization for the purposes of the voice reform rather than its normal high, medium, low income categorization? It jus- tified the choice by saying it was using the same categorization as the IMF. The stand- ard IMF classification divides countries into
‘Advanced Economies’ (AE) and Develop- ing and Emerging Economies (D&E). In the context of the 2008 quota review, the IMF made some amendments in defining its DTC category. In addition to the Developing and Emerging Economies (D&E), the Fund added six countries normally classified by the IMF as “advanced economies” and by the Bank as
“high-income”, including South Korea and Singapore. From 2008 onwards, this new set of “DTC” countries framed voice reform both in the Fund and the Bank. Table 5 shows those countries classified as high-income countries by the Bank which were incorporated in the new DTC category.
While the first set of countries result sim- ply from the fact that the World Bank and the IMF classify countries differently, the second set of countries is more problematic. Three of these countries (Czech Republic, Slovak Re- public and Slovenia) may be considered tran- sition economies, but it is difficult to see the logic underlying the decision to count South Korea, Singapore and Malta as DTCs. Four of these six countries are OECD member coun- tries (only Malta and Singapore are not).32
32 The DTC category included three further OECD mem- ber countries: Estonia, Hungary and Poland, c.f. column 1 in table 7 above.
21
According to the Bank, the rationale of using the IMF’s DTC classification instead of the LIC and MIC categories of its own statistics was to ensure comparability with voice re- forms in the IMF. Whatever the force of this rationale, it is not coincidental that the choice also had the pleasing result of making the shift of voting power achieved in the Bank appear larger than it actually was. The choice of the IMF’s DTC categorization helped the Bank move considerably closer to parity of voting power.
Before the voting power reforms of 2008 and 2010 – when the Bank was still using its own income categorizations in voice reform deliberations – achieving voting power parity
seemed a tall order; negotiating a shift from just under 35 % to 50 % would be a near-impos- sibility. In terms of the DTC category, howev- er, voting power parity was in fact very nearly achieved; the DTC share in total voting power is less than 3 percentage points away from the 50% threshold. But of the apparent major shift in favour of DTCs – 12.5 percentage points, from the pre-reform share of 34.67% to the
post-reform share of 47.19% – only 3.7% was an actual shift of voting power between coun- tries; the rest was a reclassification effect.33
making small changes appear generous
The preceding sections demonstrate that the voting power realignments were substantially more modest than they appeared to be, par- ticularly for low-income countries. Under- standably the Bank put a different gloss on it, boasting that the shift of voting power was more than two times larger than it would have been based on economic weight alone. More specifically, the Bank claimed that a phase 2
voice reform which increased the shareholding of under-represented countries so as to reflect their weight in the global economy would give DTCs only an aggregate net increase of 1.3%
Total increase: 7.19%. Shift of voting power from devel- oped to developing countries: 3.71%. Reclassification effect:
7.9-.7=.48%.
Table 5. High-income countries reclassified as “DTCs”
High-income countries classified by IMF as
Developing and Emerging Economies Countries classified by IMF as Advanced Economies, but counted as “DTC” countries in voice reform deliberations
Bahamas, Bahrain, Barbados, Brunei Darussalam, Croatia, Equatorial Guinea, Estonia, Hungary, Kuwait, Latvia, Oman, Poland, Qatar, Saudi Arabia, Trinidad &
Tobago, United Arab Emirates
Czech Republic, South Korea, Malta, Singapore, Slovak Republic, Slovenia
Table 6. The move towards parity– by different country classifications (%)
LIC+MIC DTC
NDC share before reforms 34.67 42.60
Phase 1 1.35 1.46
Phase 2 2.36 3.13
Total shift 3.71 4.59
NDC share after reforms 38.38 47.19
Note: NDC stands for “not developed countries”
22
voting power, as compared to the 3.13% ac- complished (DC 2010a: 6).34 If this 1.3 per- centage points figure is taken at face value, the phase 2 voice reform was in fact generous on the part of developed countries: they ceded more voting power to DTCs than was justified by their respective relative shares of world GDP.
But again, one should not take this argu- ment at face value. On the one hand, Bank documents on the voice reform acknowledge the pressing need to adjust IBRD shareholding to reflect the rapidly changing configuration of the global economy (as in President Zoel- lick’s historic speech of April 2010); and on the other hand they argue that such adjustment would yield only a total shift of just over 1%
of shareholding and voting power.
Unfortunately, it is not easy to penetrate much further into this matter because the Bank does not give details of how it arrived at the mysteriously low figure of 1.3 percentage points, nor of the calculations of the voting power realignment more generally. Our at- tempts to get details of the calculus, via inside sources, were vigorously rebuffed. Even Execu- tive Directors have difficulty getting them.35 But we can surmise that two factors were im- portant in determining the underestimation of voting power imbalances:
The use of a conservative benchmark for economic weight in the global economy.
Of the GDP blends considered during the Options Phase – from 60/40 at the con-
4 “An SCI that brought under-represented member’s IBRD shareholding up to their share of the global economy based on GDP 60/40, however, would yield only a .% net increase in DTC voting power for Phase 2 reforms” (DC 2010a: 6).
Technically, the Bank justifies this lack of transparency with reference to the corporate-administrative exceptions provided for under the “Access to Information” policy.
•
servative end to 30/70 at the progressive end – the most conservative option was eventually chosen.
The de facto counting of basic votes of low- income countries as over-representation of DTCs.
Thanks to these two factors the voting power realignment looked more generous than it was;
and specifically, did much less to match voting power with relative economic weight.
voting power imbalances
The overall shifts of voting power were particu- larly inadequate when seen in relation to the changing composition of world GDP. Table 7 lists the voting power to GDP ratios of the thirty largest countries (at GDP60/40). The ratio varies from below 0.5 at the bottom end (China) to almost 4 at the top end (Saudi Ara- bia).
Although basic votes and the inclusion of criteria other than GDP in the quota frame- work will by definition lead to variation in this ratio from country to country, the variation demonstrated here is far in excess of what is reasonable. Indeed, the Bank itself has typi- cally defined 0.85 as the threshold for under- representation: countries whose shareholding to economic weight ratio was below this thres- hold would then be eligible to increase their shareholding.36 But only ten out of the thirty largest countries have a voting power to GDP ratio within a 0.85 to 1.15 band of variation.
In other words, two thirds of these countries re-
This was the rule adopted, for instance, in the 998 se- lective capital increase.
•
